Waited too long for video out to pause

[Malcolm Parsons]

Binary package hint: mythtv-frontend

mythfrontend sometimes waits forever for a paused video to pause.

[mpeg2video @ 0x2af6a4e88fd0]ac-tex damaged at 0 34
[mpeg2video @ 0x2af6a4e88fd0]ac-tex damaged at 0 35
2007-04-04 18:16:28.275 NVP: prebuffering pause
2007-04-04 18:16:28.301 DPMS Deactivated
2007-04-04 18:16:29.293 Waited too long for video out to pause
2007-04-04 18:16:30.297 Waited too long for video out to pause
2007-04-04 18:16:31.301 Waited too long for video out to pause
2007-04-04 18:16:32.306 Waited too long for video out to pause
2007-04-04 18:16:33.309 Waited too long for video out to pause
2007-04-04 18:16:33.785 DPMS Deactivated
2007-04-04 18:16:34.313 Waited too long for video out to pause
2007-04-04 18:16:35.317 Waited too long for video out to pause
2007-04-04 18:16:36.321 Waited too long for video out to pause
2007-04-04 18:16:37.325 Waited too long for video out to pause
2007-04-04 18:16:38.329 Waited too long for video out to pause

...

2007-04-04 18:22:22.976 Waited too long for video out to pause
2007-04-04 18:22:23.980 Waited too long for video out to pause
2007-04-04 18:22:25.023 Waited too long for video out to pause

I'm still waiting...

[Mario Limonciello (superm1)]

Hi, thanks for the report. Has this been happening on a hidef channel, or a channel with a suboptimal signal?
I gather that there may be corruption in the stream by your top two errors.

[Malcolm Parsons (malcolm-parsons)]

It's not an HD channel.
It's a DVB-T stream, so occasional mpeg corruption is expected.
I am using a dual core processor, so bugs in threaded applications are more likely than on a single core processor.

In this case it looks like one thread is waiting to be signalled that another thread has paused the video, but the other thread has paused the video and notified all waiting threads before the first thread started to wait.

[David Morris (dave-greenacre)]

"I am using a dual core processor, so bugs in threaded applications are more likely than on a single core processor.

In this case it looks like one thread is waiting to be signalled that another thread has paused the video, but the other thread has paused the video and notified all waiting threads before the first thread started to wait."

are you able to disable one core to confirm this?

[Malcolm Parsons (malcolm-parsons)]

I can try.

I can reproduce this bug by randomly pressing the FF, Rev, Play and Pause buttons on my IR remote for 30 seconds whilst playing a recording.

[Malcolm Parsons (malcolm-parsons)]

If I boot with "nosmp" on the kernel command line, then the boot fails because interrupts from the hard drives are not received.

So I can't test whether mythtv is more stable on a single core CPU.

[David Morris (dave-greenacre)]

Look for package "schedutils" in synaptic - one of the tools is "taskset" for setting the CPU a process can run on. Tomorrow at work I'll see how I can turn one of my cores off

[Malcolm Parsons (malcolm-parsons)]

I can reproduce this bug with mythfrontend forced to run on 1 CPU by using taskset

[David Morris (dave-greenacre)]

This has been fixed in the last few weeks, and need packaging now.

[radiojammer3699 (radiojammer3699)]

Hi,
So does this thread mean that a fix for this bug is coming to ubuntu repos soon? a fix for this would really make my day ;)

I have suffered about this same bug since I installed mythtv on ubuntu 6.10 early this year (my first "working" install of ubuntu+mythtv). I use the mythtv packages from ubuntu repos. I have dualcore AMD A64 x2 with power ondemand set (limited to 1.8GHz...2.4GHz) on my mythtv-box. I am planing to upgrade to ubuntu 7.04 someday...

[Mario Limonciello (superm1)]

radiojammer3699 wrote:
> Hi,
> So does this thread mean that a fix for this bug is coming to ubuntu repos soon? a fix for this would really make my day ;)
>
> I have suffered about this same bug since I installed mythtv on ubuntu
> 6.10 early this year (my first "working" install of ubuntu+mythtv). I
> use the mythtv packages from ubuntu repos. I have dualcore AMD A64 x2
> with power ondemand set (limited to 1.8GHz...2.4GHz) on my mythtv-box. I
> am planing to upgrade to ubuntu 7.04 someday...
>

This means that the current development release (Gutsy) has the fixes
applied. There has been significant packaging changes that would like
prevent a backport to edgy (and possibly feisty). If you do move up to
feisty however, there are weekly builds available that will likely have
this fix applied soon available at mythbuntu.org/auto-builds

[Ubuntu QA Website (ubuntuqa)]

This bug has been reported on the Ubuntu Package testing tracker.

A list of all reports related to this bug can be found here:
http://packages.qa.ubuntu.com/qatracker/reports/bugs/102964